Reactivity of aldehydes with semi-stabilised arsonium ylide anions: Synthesis of terminal (E)-l,3-dienes

Article abstract of DOI:10.1002/chem.200601549

A study of the reactivity of semi-stabilised arsonium ylide anions in olefinalion reactions is presented. The different ylide anions were generated by the addition of nBuLi to various arsonium halide derivatives: [Ph ₂As(R)R']⁺X , where R and R′ are methyl, allyl, prenyl or benzyl groups. By using diallyldiphenylarsonium bromide (R = R′ = allyl) an olefinationprotocol was optimised allowing the efficient transformation of aliphatic aldehydes into terminal 1,3-dienes with a high selectivity for the E isomer (E/Z ratios ranging from 90:10 to 97:3). The olefination reactions of aldehydes with dissymmetric arsonium halides (R≠R′) are very chemoselective; with arsonium ylide anions the benzyl moiety is more reactive than the allyl moiety which is much more reactive than prenyl and methyl groups. Based on the experimental results, a mechanism is proposed for the reaction.

Full text of DOI:10.1002/chem.200601549

FULL PAPER
DOI: 10.1002/chem.200601549
Reactivity of Aldehydes with Semi-Stabilised Arsonium Ylide Anions:
Synthesis of Terminal (E)-1,3-Dienes
Damien Habrant, Bruno Stengel, StØphane Meunier,* and Charles Mioskowski*[a]
Abstract: A study of the reactivity of
semi-stabilised arsonium ylide anions
in olefination reactions is presented.
The different ylide anions were gener-
ated by the addition of nBuLi to vari-
ous arsonium halide derivatives:
[Ph2As(R)R’]+Xꢀ, where R and R’ are
methyl, allyl, prenyl or benzyl groups.
By using diallyldiphenylarsonium bro-
mide (R=R’=allyl) an olefination
protocol was optimised allowing the ef-
ficient transformation of aliphatic alde-
hydes into terminal 1,3-dienes with a
high selectivity for the E isomer (E/Z
ratios ranging from 90:10 to 97:3). The
olefination reactions of aldehydes with
dissymmetric arsonium halides (R¼R’)
are very chemoselective; with arsonium
ylide anions the benzyl moiety is more
reactive than the allyl moiety which is
much more reactive than prenyl and
methyl groups. Based on the experi-
mental results, a mechanism is pro-
posed for the reaction.
Keywords: 1,3-dienes
· alkenes ·
arsenic · olefination · ylide anions ·
ylides
Introduction
by using ortho-substituted triphenylphosphorus ylides, albeit
in low-to-moderate yields as a result of steric congestion
(12%, E/Z 2:98, from heptanal).[7]
The Wittig reaction, discovered in 1953,[1] is widely recog-
nised as a good method for the olefination of aldehydes and
ketones.[2] The stereochemistry of the olefination reaction is
strongly dependent on the type of ylide and the exact reac-
tion conditions. Reactive and stabilised phosphorus ylides
can be structurally tuned to give either (Z) or (E)-olefins
with high selectivity.[3] However, semi-stabilised (or moder-
ately reactive) phosphorus ylides, bearing mildly conjugating
substituents (a-1-alkenyl, a-1-alkynyl, a-1-aryl, a-1-hetero-
aryl, a-1-halo, a-1-alkoxy) often show no great stereoselec-
tive preference in the olefination of carbonyl compounds.[4]
Efficient protocols for the construction of unsubstituted ter-
minal (E)-1,3-dienes (R-CH=CHCH=CH2) with semi-stabi-
lised phosphorous ylides involve the use of a-deprotonated
allylic diphenylphosphine oxides[5] and phosphonates.[6] With
aliphatic aldehydes, these reagents lead to the formation of
dienes having E/Z ratios in the range of 99:1 to 94:6 in mod-
erate-to-good yields (45–78%). High cis selectivity in the
synthesis of unsubstituted terminal 1,3-dienes was achieved
Arsonium ylides are interesting reagents for carbonyl ole-
fination reactions as they are reported to be stronger nucle-
ophiles than the corresponding phosphonium ylides. The in-
creased negative charge density at the carbon centre of arso-
nium ylides as compared with phosphonium ylides accounts
for the difference in reactivity observed.[8] It is known that
the reactions of semi-stabilised arsonium ylides with carbon-
yl compounds result in a mixture of olefin and epoxide
products.[9] Hsi and Koreeda have demonstrated that the use
of LiHMDS or KHMDS in the generation of the ylide from
allyltriphenylarsonium tetrafluoroborate directs the reaction
towards the formation of epoxides or olefins, respectively.[10]
By using KHMDS, unsubstituted terminal (E)-1,3-dienes
were obtained in moderate-to-good yields from hindered ali-
phatic and aromatic aldehydes with no traces of the corre-
sponding (Z)-1,3-dienes. In an attempt to improve the selec-
tivity of the reaction of arsonium semi-stabilised ylides with
aldehydes, our groupreported the synthesis of the first arso-
nium ylide anion. The dibenzyl ylide anion obtained reacts
with hexanal in THF/HMPA (5:1) and leads exclusively to
alkene formation with an E/Z ratio of >99:1.[11] Similar re-
sults were obtained with benzyl(2-hydroxyethyl)arsonium
ylide anions in the presence of HMPA.[12] The dibenzylarso-
nium ylide anion converts only one equivalent of aldehyde
to the olefination product, whereas a similar phosphonium
ylide anion converts two equivalents of aldehyde.[13] It was
[a] D. Habrant, Dr. B. Stengel, Dr. S. Meunier, Dr. C. Mioskowski
Laboratoire de Synth›se Bio-Organique, UMR 7175/LC1-CNRS
FacultØ de Pharmacie, UniversitØ Louis Pasteur
74 Route du Rhin, BP 24, 67401 Illkirch (France)
Fax : (+33)390-244-306
E-mail: meunier@bioorga.u-strasbg.fr
mioskow@aspirine.u-strasbg.fr
Chem. Eur. J. 2007, 13, 5433 – 5440
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5433

suggested that the arsonium oxide anion, produced by olefi-
nation of the first equivalent of aldehyde, reacts with a
second equivalent of the aldehyde, but that the b-hydroxyar-
sine oxide intermediate formed does not react further to
form an olefin.[11] This hypothesis was supported by the ob-
first step, lithium diphenylarsine, which is prepared in situ
by treatment of triphenylarsine with lithium in THF, is treat-
ed with allyl bromide, to afford allyldiphenylarsine (2).
During the reaction, the addition of one equivalent of tert-
butyl chloride is necessary to quench the phenyllithium
formed as a side product. Compound 2 is purified by distilla-
tion under vacuum and should be kept under an inert at-
mosphere. In the second step, 2 is quaternised with an halo-
genated compound in acetonitrile to give 1a–d in good
yields.
servation that lithiomethyl(diphenyl)arsine oxide reacts with
A
aldehydes to give a stable b-hydroxyarsine oxide deriva-
tive.[14]
In the course of our studies it became apparent that an ef-
ficient method for the synthesis of a variety of unsubstituted
terminal (E)-1,3-dienes was needed. In our exploration of
new synthetic methodologies we investigated the use of ar-
sonium allylic ylide anions. Herein is presented the synthesis
of diallyldiphenylarsonium bromide 1a and the study of the
reactivity of the corresponding ylide anion in the presence
of aldehydes (Scheme 1). The results show that terminal
Reactivity of diallyldiphenylarsonium bromide 1a with alde-
hydes: Compound 1a was the first arsonium bromide that
we synthesised with the aim of studying a new methodology
for the synthesis of unsubstituted terminal (E)-1,3-dienes
(R-CH=CHCH=CH2) from aldehydes (Scheme 3).
Scheme 1. Structure of diphenylarsonium halide derivatives 1a–d.
Scheme 3. Synthesis of (E)-1,3-dienes from aldehydes and the ylide anion
generated from 1a.
(E)-1,3-dienes are obtained in good yields and with high se-
lectivities (E/Z 90:10 to 97:3). Three other diphenylarsoni-
um halides derivatives 1b–d have also been synthesised and
the reactivity of the corresponding ylide anions in olefina-
tion reactions studied (Scheme 1). The results show that 1b
and 1c exclusively transfer the allyl chain to the aldehydes,
whereas 1d transfers the benzyl chain with high selectivity.
Based on these observations and complementary deuteria-
tion experiments, a mechanism is proposed for the reac-
tion.
Hexadecanal was used to optimise the experimental con-
ditions. The treatment of one equivalent of 1a with two
equivalents of nBuLi and subsequent addition of the alde-
hyde led to the expected diene with an E/Z ratio of 90:10
Table 1. Optimisation of the conditions for the reactions between diphe-
nylarsonium halide 1a and hexadecanal.[a]
Entry
nBuLi
[equiv]
Hexadecanal
[equiv]
Yield[b] [%]
E/Z ratio
A
ACHTREUNG
1
2
3
2
3
3.5
1
3
3.5
40
76
73
90:10
94:6
92:8
Results and Discussion
[a] Diphenylarsonium halide 1a was treated with nBuLi in THF/HMPT
(5:1) at ꢀ358C for 3 h before addition of hexadecanal at ꢀ788C. After
24 h at room temperature and hydrolysis, nonadeca-1,3-diene was ex-
tracted and purified as an inseparable mixture of E and Z isomers.
[b] Yields are based on 1a.
Synthesis of diphenylarsonium halide derivatives: Com-
pounds 1a–d were synthesised by following the strategy de-
scribed by our groupfor the formation of dibenzyldipheny-
larsonium bromide.[11] Starting from commercially available
triphenylarsine, this strategy allows the synthesis in two
steps of diphenylarsonium halide derivatives bearing two
different substituents: [Ph2As(R)R’]+Xꢀ, where R and R’
are methyl, allyl, prenyl or benzyl groups (Scheme 2). In the
(Table 1, entry 1). The E/Z ratio and yield increased when
three equivalents of nBuLi and aldehyde were used; nona-
deca-1,3-diene was obtained in 76% yield and with a 94:6
E/Z ratio (Table 1, entry 2). Larger quantities of reactants
led to a decrease in both yield and stereoselectivity (Table 1,
entry 3). As described for other arsonium ylide anions, no
reaction was observed in the absence of HMPT. Yields
never exceeded 100% suggesting that 1a can convert only
one equivalent of aldehyde to the diene product; a similar
reactivity was observed with the arsonium ylide anion pro-
duced from dibenzyldiphenylarsonium bromide.[11]
Scheme 2. Synthesis of diphenylarsonium halide derivatives 1a–d. Re-
agents and conditions: a) 1) Li, THF, 12 h at RT; 2) tBuCl, THF, 2 h,
08C!RT; 3) allyl bromide, THF, 18 h, 08C!RT; b) RX, CH3CN, 24 h at
reflux for 1a,b, at RT for 1c,d.
5434
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 5433 – 5440

Reactivity of Aldehydes with Arsonium Ylide Anions
FULL PAPER
The quantitative formation of the arsonium ylide anion
under the optimised experimental conditions was confirmed
by deuteriation. The dark red solution resulting from the
treatment of 1a with three equivalents of nBuLi was
quenched by the addition of DCl in D2O; the 1H NMR spec-
trum of the arsonium salt thus obtained (with a yield of
90% after purification) showed the insertion of one deuteri-
um atom into the a position of each of the allylic groups of
1a.
The reactivity of 1a with other aldehydes was then stud-
ied. Hexadecanal and dodecanal reacted similarly, giving the
terminal (E)-1,3-dienes 4a and 4b in good yields and with
very good E/Z ratios (Table 2, entries 1 and 2). Steric hin-
olefination reactions, which can be the case when using
semi-stabilised arsonium ylides.
Reactivity of diphenylarsonium halide derivatives 1b–d:
The results presented above suggest that diallyldiphenylar-
sonium bromide 1a, in the presence of nBuLi, olefinates
1 equivalent of aldehyde. Therefore, only one of the two al-
lylic groups on 1a is transferred during the process; the
second one does not give an olefination product. Arsonium
halides 1b–d bearing different substituents were then syn-
thesised to study the chemoselectivity of their olefination re-
actions with aldehydes.
Following the experimental protocol optimised for 1a, ar-
sonium halides 1b–d were treated with 3 equivalents of
nBuLi and 3 equivalents of aldehydes (Table 3). Arsonium
halides 1b and 1c reacted with excellent selectivities as only
unsubstituted terminal 1,3-dienes (R-CH=CHCH=CH2)
were formed (Table 3, entries 1–5). These results demon-
strate that aldehydes react with a very high selectivity with
the deprotonated allyl chain of the ylide anions obtained
from 1b and 1c. The deprotonated prenyl chain of the 1b
ylide anion and the deprotonated methyl group in the 1c
ylide anion do not produce olefination adducts. With 1b,c
the stereoselectivities of the olefination reactions are very
high (E/Z>98:2). Diene 4 f was obtained from dihydrocin-
namaldehyde and bifunctional ylide anions 1b,c with E/Z
ratios higher than that obtained with the diallyl ylide anion
1a (compare Table 2, entry 6 and Table 3, entries 2 and 5).
In the olefination of dodecanal and dihydrocinnamaldehyde
with 1d the selectivity is lower as two different alkenes are
isolated (Table 3, entries 6 and 7). The major products 7b
and 7 f result from the transfer of the deprotonated benzyl
chain of the ylide anion 1d to the aldehydes; the side-prod-
ucts 4b and 4 f result from the transfer of the deprotonated
allyl chain. In each case the (E)-alkene was formed with
high stereoselectivity. Interestingly, with benzaldehyde,
trans-stilbene 7g was obtained as the only reaction product
(Table 3, entry 8). This reflects the unreactivity of benzalde-
hyde with the ylide anion obtained from 1a and the higher
reactivity of the benzyl chain of 1d relative to its allyl chain.
Similarly, olefination of trans-cinnamaldehyde in the pres-
ence of 1d led to diphenylbutadiene 7h as the only reaction
product (Table 3, entry 9). Products 7b and 7f–h were ob-
tained as pure E isomers; this stereospecificity was previous-
ly observed with dibenzyldiphenylarsonium bromides.[11]
To conclude, our results highlight the following hierarchy
in the reactivity of the ylide anion side chains with alde-
hydes: benzyl>allyl@prenyl and methyl.
Table 2. Reactions between diphenylarsonium halide 1a and aldehydes.[a]
Entry
Aldehyde 3
Product 4
Yield[b] E/Z
[%]
ratio
1
2
76
94:6
79
72
94:6
97:3
3
4
5
41
85
93:7
94:6
6
7
8
64
0
90:10
–
–
–
–
0
[a] Diphenylarsonium halide 1a was treated with nBuLi (3 equiv) in
THF/HMPT (5:1) at ꢀ358C for 3 h before addition of the aldehyde
(3 equiv) at ꢀ788C. The reactions were stopped after 24 h at room tem-
perature. [b] Yields correspond to purified compounds and are based on
1a.
drance at the a or b position of the aldehyde leads to a de-
crease in the reaction yield, but not in the stereoselectivity
(Table 2, entries 3 and 4). Furthermore, the reaction worked
equally well in the presence of a protected 1,2-diol (Table 2,
entry 5). Reaction of 1a with dihydrocinnamaldehyde led to
the expected terminal 1,3-diene 4 f, albeit with a lower ste-
reoselectivity (E/Z 90:10, entry 6). Finally, the use of benzal-
dehyde as well as trans-cinnamaldehyde led to complex mix-
tures from which the expected products could not be puri-
fied (Table 2, entries 7 and 8). These results are similar to
the best results obtained with lithiated allylic phosphonates
in terms of yield and selectivity.[5] The stereoselectivities are
slightly lower than those obtained with a-deprotonated allyl-
ic diphenylphosphine oxides, although they lead to better
yields.[6] No traces of epoxide products were detected in our
Mechanistic hypotheses: Herein we present our hypotheses
for the rationalisation of the reactivity of the semi-stabilised
arsonium ylide anions 1a–d. The selectivities observed for
the reactions of aldehydes with the ylide anions generated
from 1b–d was not expected. Indeed, in dianionic systems,
electrophiles are expected to react with the less stable
anionic site.[15] By considering the ylide anions generated
from 1b–d as dianionic systems, it is reasonable to state that
Chem. Eur. J. 2007, 13, 5433 – 5440
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5435

S. Meunier, C. Mioskowski et al.
Table 3. Reactions between diphenylarsonium halides 1b–d and aldehydes.[a]
sine oxide intermediate in the
reaction media. When 1a or 1b
was used the only arsenic-based
reaction adducts that we could
detect were allyldiphenylarsine
oxide and prenyldiphenylarsine
oxide, respectively, which were
not expected on the basis of the
initial mechanistic hypothesis.
These observations suggest that
the mechanism involved in the
reactions of semi-stabilised ar-
sonium ylide anions 1a–d with
aldehydes is different to the
one involved in the reaction of
the corresponding phosphonium
species.
For reasons of clarity our
mechanistic hypothesis will be
illustrated with the arsonium
bromide 1b (Scheme 4). The in-
itial attack of the arsonium
ylide anion on the carbonyl
carbon atom of the aldehyde
produces oxaarsetane I or II.
Oxaarsetane I is favoured ki-
netically and thermodynamical-
ly because the residual anionic
charge in I is stabilised more
than the residual anionic charge
in II. Intermediate I should lead
to the olefination adduct 5, but
this adduct was not observed
experimentally. To rationalise
our results we suggest that ox-
aarsetane I can revert back to
the ylide anion and aldehyde
Entry
Aldehyde
Arsonium halide
Products, yield[b] (E/Z ratio)
1
1b
2
3
1b
1b
4
5
6
1c
1c
1d
7
8
9
1d
1d
1d
[a] Diphenylarsonium halides 1b–d were treated with nBuLi (3 equiv) in THF/HMPT (5:1) at ꢀ358C for 3 h
before addition of the aldehyde (3 equiv) at ꢀ788C. The reactions were stopped after 24 h at room tempera-
ture. [b] Yields correspond to purified compounds (0% is indicated if the compound was not formed in the re-
action) and are based on 1b–d.
1) for 1b, the allyl anion is more stable than the prenyl
anion (electron-donating effects of the two methyl groups),
2) for 1c, the allyl anion is more stable than the methyl
anion and 3) for 1d, the benzyl anion is more stable than
the allyl anion (electron delocalisation onto the phenyl
ring). In each case the major olefination adduct observed re-
sults from the reaction of the aldehyde with the more stable
anionic site of the ylide anion.
Our first mechanistic hypothesis was based on the reactiv-
ities of the phosphonium ylide anions[13] and on the previous
study of the dibenzylarsonium ylide anion.[11] As presented
in the introduction of this report, it is generally accepted
that in both cases a first equivalent of aldehyde reacts with
the ylide anion and leads to an olefination adduct and to a
phosphonium or arsonium oxide anion. These intermediates
then react with a second equivalent of aldehyde to give a b-
hydroxyphosphine oxide or a b-hydroxyarsine oxide. In the
latter case, the intermediate is expected to be too stable to
evolve further and generate an olefin adduct.[14] Despite sev-
eral attempts we could not trap or observe a b-hydroxyar-
and that II might evolve into a more stable intermediate.
We suggest that oxaarsetanes I and II are in equilibrium
with the intermediates III and IV through an intramolecular
transmetallation reaction. This process is facilitated thanks
to the high acidity of the allylic protons in semi-stabilised
ylide anions.[16] Such an equilibrium was presented by
McKenna and Walker in order to rationalise the high E/Z
ratios obtained in the reactions of aldehydes with semi-stabi-
lised phosphonium ylide anions (an equilibrium forms be-
tween cis- and trans-oxaphosphetane anions).[13]
Intermediate IV is expected to be more stable than III;
the anionic charge on III is destabilised by the electron-do-
nating effects of the two methyl groups of the prenyl chain.
Intermediate IV would naturally lead, after protonation, to
the unsubstituted terminal 1,3-diene 4 and to prenyldipheny-
larsine oxide, which was detected as a major side product in
our reaction mixtures. According to this mechanism the for-
mation of the arsine oxide from IV prevents the olefination
of a second equivalent of aldehyde.
5436
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 5433 – 5440

Reactivity of Aldehydes with Arsonium Ylide Anions
FULL PAPER
Scheme 4. Mechanism proposed for aldehyde olefination by the ylide anion generated from 1b.
Concerning the stereoselectivity of the olefination pro-
cess, intermediates III and IV can be compared with the b-
oxidophosphonium ylides obtained in SCOOPY-type reac-
tions (a-substitution plus carbonyl olefination via b-oxido
phosphorus ylides).[17] In this procedure a b-oxidophosphoni-
um, resulting from the addition of a phosphonium ylide to
an aldehyde, is deprotonated by a strong base to give a
stable b-oxidophosphonium ylide which is subsequently
trapped by an electrophile (Scheme 5). If a proton donor is
bottom). This result is indirect, but strong proof of the for-
mation of intermediate IV in the reaction process
(Scheme 4). Other similar deuteriation experiments were
performed, but with longer reaction times at ꢀ788C or
room temperature using a variety of deuterium donors
(D2O, DCl, CD3OD). These experiments gave 4 f in higher
yields, but with lower deuterium insertion, which suggests
that intermediate IV was neutralised in the reaction mix-
ture. We have not yet determined the mechanism for the in
situ neutralisation of intermediate IV; the use of [D8]THF
Scheme 5. Illustration of the SCOOPY-type mechanism for the synthesis
of (E)-olefins from a phosphonium ylide and an aldehyde.
used, the b-oxidophosphonium ylide is capable of stereospe-
cific protonation to form the threo betaine which is the pre-
cursor of the (E)-olefin. Similarly, it can be expected that
protonation of III or IV would lead to the (E)-olefin with
high stereoselectivity, which rationalises our experimental
observations.
To validate our mechanistic hypotheses, we made several
attempts to trap intermediate IV with a deuterium donor.
The best result was obtained when arsonium bromide 1b
was treated at ꢀ358C with nBuLi (3 equiv) in THF for 3 h,
followed by the addition of dihydrocinnamaldehyde
(3 equiv) at ꢀ788C. After 2 h at ꢀ788C, the solution was
quenched with deuteriated methanol. After column chroma-
tography, the expected (E)-1,3-diene 4 f was obtained with a
yield of 21%. 1H NMR analysis showed the insertion of
deuterium in the expected position of the 1,3-diene moiety
of 4 f with a 64% deuterium/hydrogen ratio (Figure 1,
Figure 1. 1H NMR spectra of hydrogenated (top) and deuteriated
(bottom) 4 f (E/Z 92:8).
Chem. Eur. J. 2007, 13, 5433 – 5440
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5437

S. Meunier, C. Mioskowski et al.
0.34 mol) in THF (40 mL) was added dropwise. The pale yellow solution
obtained was stirred for 18 h at room temperature. The mixture was then
diluted with diethyl ether and hydrolyzed with water. The organic extract
was washed with brine, dried with magnesium sulfate and concentrated
under vacuum. The crude product was then distilled under reduced pres-
sure to afford 2 (29.06 g, 66%) as a colourless oil. B.p. 109–1138C/
0.1 mmHg. 1H NMR (200 MHZ, CDCl3): d=7.50–7.30 (m, 10H), 6.0–
5.75 (m, 1H), 5.04–4.94 (m, 2H), 2.87 ppm (d, J=7.9 Hz, 1H); 13C NMR
(50 MHz, CDCl3): d=140.2, 134.0, 133.0, 128.4, 116.1, 32.8 ppm; IR
did not lead to the deuteriation of compound 4, the excess
of aldehyde might be responsible for the protonation of IV.
In the optimised experimental procedure three equiva-
lents of nBuLi are used; the excess nBuLi probably reacts
with the excess aldehyde although it could also participate
in the deprotonation of intermediate II (as in a SCOOPY-
type reaction) leading to a b-oxidoarsonium ylide anion. We
consider this chemical pathway unlikely because, in the vari-
ous deuteriation experiments, no deuterium atom insertion
was detected in the prenylarsine oxide produced.
(CsI): n˜ =3068, 3052, 2971, 2911, 1581, 1481, 1434, 1075, 735 cmꢀ1
Diallyldiphenylarsonium bromide (1a): solution of 2 (1.23 g,
.
A
4.55 mmol) and allyl bromide (1.14 mL, 13.20 mmol) in acetonitrile
(2.5 mL) was refluxed for 24 h. The solvent was removed under vacuum
and the solid obtained was dissolved in dry DCM. The reaction product
was precipitated by addition of dry diethyl ether to afford 1a (1.74 g,
Conclusion
1
98%) as white crystals. F.p. 142–1448C. H NMR (200 MHz, CDCl3): d=
7.83–7.60 (m, 10H), 5.88–5.65 (m, 1H), 5.45 (d, J=16.9 Hz, 1H), 5.26 (d,
J=9.9 Hz, 1H), 4.44 ppm (d, J=7.6 Hz, 2H); 13C NMR (50 MHz,
CDCl3): d=133.6, 132.9, 130.3, 124.7, 124.6, 120.2, 113.9, 29.8 ppm; IR
(KBr): n˜ =3054, 3010, 2934, 2881, 2799, 1634, 1436, 1405, 1202, 1084, 990,
743 cmꢀ1; elemental analysis calcd (%) for C18H20AsBr (391.18): C 55.27,
H 5.15, As 19.15, Br 20.43; found: C 55.21, H 5.21, As 18.65, Br 20.45.
In conclusion, we have proposed a new protocol for the syn-
thesis of unsubstituted terminal (E)-1,3-dienes from alde-
hydes using semi-stabilised arsonium ylide anions. The ylide
anion was generated from diallyldiphenylarsonium bromide
1a and allowed the olefination of one equivalent of alde-
hyde. Starting from various aliphatic aldehydes the dienes
were obtained in good yields and with high stereoselectivi-
ties (E/Z ratios ranging from 90:10 to 97:3). The study of ar-
sonium semi-stabilised ylide anions bearing two different
anionic substituents (methyl, allyl, prenyl or benzyl) allowed
us to propose a mechanism for the olefination reaction. The
key stepof this mechanism is an intramolecular transmetal-
lation process leading to intermediate IV which is closely re-
lated to the b-oxidophosphonium ylides observed in
SCOOPY-type reactions. Stereospecific protonation of IV
leads to the formation of an E olefination adduct and an
arsine oxide. The mechanism proposed for this reaction is
different to the mechanism reported for the olefination of
carbonyl compounds by phosphonium ylide anions.
(3-Methylbut-2-ene)(prop-2-enyl)diphenylarsonium bromide (1b): A so-
lution of 2 (1.21 g, 4.47 mmol) and 1-bromo-3-methylbut-2-ene (1.51 mL,
12.98 mmol) in acetonitrile (2.5 mL) was refluxed for 24 h. The solvent
was removed under vacuum and the solid obtained was dissolved in dry
DCM. The reaction product was precipitated by addition of dry diethyl
ether to afford 1b (1.12 g, 60%) as white crystals. F.p. 122–1248C. 1H
NMR (200 MHz, CDCl3): d=7.80–7.54 (m, 10H), 5.83–5.65 (m, 1H),
5.45 (d, J=16.9 Hz, 1H), 5.28 (d, J=9.9 Hz, 1H), 5.14 (t, J=8.6 Hz, 1H),
4.47 (d, J=7.7 Hz, 2H), 4.40 (d, J=8.6 Hz, 2H), 1.62 (s, 3H), 1.37 ppm
(s, 3H); 13C NMR (50 MHz, CDCl3): d=142.6, 133.0, 132.6, 129.8, 124.5,
123.9, 120.0, 109.2, 29.5, 25.6, 25.2, 18.0 ppm; IR (KBr): n˜ =3016, 2931,
2903, 1654, 1636, 1438, 938, 746 cmꢀ1; elemental analysis calcd (%) for
C20H24AsBr (419.23): C 57.30, H 5.77, As 17.87, Br 19.06; found: C 57.01,
H 5.93, As 17.78, Br 18.88.
Methyl(prop-2-enyl)diphenylarsonium iodide (1c):
A solution of 2
(1.19 g; 4.44 mmol) and methyl iodide (0.80 mL, 12.86 mmol) in acetoni-
trile (2.5 mL) was stirred at room temperature for 24 h. The solvent was
removed under vacuum and the solid obtained was dissolved in dry
DCM. The reaction product was precipitated by addition of dry diethyl
ether to afford 1c (1.73 g, 95%) as white crystals. F.p. 141–1428C. 1H
NMR (200 MHz, CDCl3): d=7.87–7.58 (m, 10H), 5.89–5.71 (m, 1H),
5.64 (d, J=16.8 Hz, 1H), 5.35 (d, J=9.8 Hz, 1H), 4.35 (d, J=3.3 Hz,
2H), 4.40 (d, J=8.6 Hz, 2H), 2.79 ppm (s, 3H); 13C NMR (50 MHz,
CDCl3): d=133.6, 132.1, 130.4, 125.0, 124.2, 121.5, 30.4, 7.7 ppm; IR
(KBr): n˜ =3054, 3018, 2971, 2936, 2895, 1632, 1438, 1338, 1201, 997,
747 cmꢀ1; elemental analysis calcd (%) for C16H18AsI (412.14): C 46.63,
H 4.40, As 18.18, I 30.79; found: C 46.56, H 4.56, As 18.00, I 30.70.
Experimental Section
General methods: All experiments were carried out under argon. Com-
mercially available aldehydes were freshly distilled prior to use. THF and
diethyl ether were distilled from sodium and benzophenone. HMPT was
distilled over calcium hydride and stored over 4 Š molecular sieves
under argon. Acetonitrile and DCM were distilled over sodium hydride.
TLC was performed on Merck silica gel 60F54 and were detected by
using UV light at 254 nm and vanilline. Silica gel (Merck 60, 40–63 mm)
was used for flash column chromatography. NMR spectra were recorded
at 200 or 300 MHz for 1H NMR and 50 or 75 MHz for 13C NMR, using
chloroform as the internal reference (7.26 ppm for 1H and 77.16 ppm for
13C). Infra-red spectra (KBr discs or CsI films) were recorded on a
Perkin-Elmer apparatus (1600 FT-IR). Fusion points were recorded on a
Reichert-Jung (Thermo Galen) apparatus. Elemental analyses were per-
formed by the Service Central d’Analyses du CNRS at Vernaison
(France). The E/Z ratios were calculated from the integration values of
(Benzyl)(prop-2-enyl)diphenylarsonium bromide (1d): A solution of 2
(1.26 g, 4.68 mmol) and benzyl bromide (1.61 mL, 13.57 mmol) in aceto-
nitrile (2.5 mL) was stirred at room temperature for 24 h. The solvent
was removed under vacuum and the solid obtained was dissolved in dry
DCM. The reaction product was precipitated by addition of dry diethyl
ether to afford 1d (1.82 g, 88%) as white crystals. F.p. 159–1618C. 1H
NMR (200 MHz, CDCl3): d=7.70–7.46 (m, 10H), 7.27–7.12 (m, 5H),
5.78–5.59 (m, 1H), 5.38 (d, J=16.8 Hz, 1H), 5.24 (d, J=9.9 Hz, 1H), 5.09
(s, 2H), 4.41 ppm (d, J=7.6 Hz, 2H); 13C NMR (50 MHz, CDCl3): d=
133.1, 132.8, 129.8, 128.7, 128.3, 127.7, 124.3, 124.0, 119.7, 32.0, 28.9 ppm;
IR (KBr): n˜ =3048, 3017, 2932, 2880, 2802, 1636, 1605, 1492, 1323, 995,
744 cmꢀ1; elemental analysis calcd (%) for C22H22AsBr (441.23): C 59.89,
H 5.03, As 16.98, Br 18.11; found: C 59.25, H 5.04, As 16.30, Br 17.95.
1
the H NMR spectra.
(Prop-2-enyl)diphenylarsine (2): Lithium (3.4 g, 0.49 mol) was added to a
solution of triphenylarsine (50 g, 0.16 mol) in THF (250 mL). After a few
minutes, the solution became dark red and stirring was maintained for
24 h. Excess lithium was then removed by cannulating the solution into
another flask under argon. A solution of tert-butyl chloride (18 mL,
0.16 mol) in THF (40 mL) was added dropwise at 08C to destroy the phe-
nyllithium formed during the reaction. The mixture was stirred for 2 h at
room temperature and became pale red. Then allyl bromide (29.7 mL,
General procedure for the olefination reaction: The arsonium salt 1a
(0.50 g, 1.28 mmol) was suspended in THF (25 mL) and HMPT (5 mL)
ꢀ358C.[18] Then a solution of n-butyllithium (1.6m in hexanes, 2.55 mL,
4.08 mmol)[19] was added dropwise and the dark red mixture was stirred
at ꢀ358C for 3 h. The solution was then cooled to ꢀ788C and a solution
of aldehyde (4.08 mmol) in THF (9 mL) was added dropwise. After
5438
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 5433 – 5440

Reactivity of Aldehydes with Arsonium Ylide Anions
FULL PAPER
10 min at ꢀ788C, the solution was allowed to warm to room temperature
and stirring was continued for 24 h. Diethyl ether was then added and
the reaction mixture was hydrolyzed with water. The organic extracts
were washed with brine, dried with magnesium sulfate and concentrated
under vacuum. The crude mixture was then purified on silica gel using
cyclohexane/diethyl ether (95:5) as eluent.
(3E)-Hexa-3,5-dienylbenzene (4 f): Compound 4 f was obtained from 1a
and 3-phenylpropanal (3 f) in 64% yield as a mixture of isomers (E/Z
90:10) following the general olefination procedure. Comparison of the
spectral data with the literature[23] confirmed the identity of compound
4 f. 1H NMR (300 MHz, CDCl3): d=7.43–7.23 (m, 5H), 6.83–6.63 (m,
0.10H), 6.41 (ddd, J=16.8, 11.6, 10.2 Hz, 0.90H), 6.19 (dd, J=15.0,
11.6 Hz, 1H), 5.84 (dt, J=15.0, 6.7 Hz, 0.90H), 5.67–5.43 (m, 0.10H),
5.20 (dd, J=16.8, 1.7 Hz, 1H), 5.07 (dd, J=10.2, 1.7 Hz, 1H), 2.82 (t, J=
9.7 Hz, 2H), 2.68–2.43 ppm (m, 2H); 13C NMR (75 MHz, CDCl3): d=
141.8, 137.2, 134.2, 132.2, 131.5, 130.8, 128.4, 128.3, 125.9, 117.1, 115.0,
35.9, 35.7, 34.3, 29.5 ppm; IR (CsI): n˜ =3061, 3026, 2927, 1710, 1602,
(E)-Nonadeca-1,3-diene (4a): Compound 4a was obtained from 1a and
palmitaldehyde (3a) in 76% yield as a mixture of isomers (E/Z 94:6) fol-
lowing the general olefination procedure. 1H NMR (200 MHz, CDCl3):
d=6.76–6.56 (m, 0.06H), 6.32 (ddd, J=17.0, 10.2, 10.1 Hz, 0.94H), 6.06
(dd, J=15.2, 10.2 Hz, 1H), 5.71 (dt, J=15.2, 6.9 Hz, 0.94H), 5.37–5.59
(m, 0.06H), 5.09 (dd, J=17.0, 1.8 Hz, 1H), 4.96 (dd, J=10.1, 1.8 Hz,
1H), 2.30–2.03 (m, 2H), 1.50–1.10 (m, 26H), 0.87 ppm (t, J=6.5 Hz,
3H); 13C NMR (50 MHz, CDCl3): d=137.4, 135.6, 132.4, 130.8, 116.6,
114.5, 32.6, 31.9, 29.7, 29.6, 29.5, 29.4, 29.2, 27.8, 22.7, 14.1 ppm; IR (CsI):
n˜ =3086, 2925, 2854, 1795, 1652, 1603, 1464, 1377, 1001, 949, 722 cmꢀ1; el-
emental analysis calcd (%) for C19H36 (264.50): C 86.28, H 13.72; found:
C 86.15, H 13.70.
1495, 1089, 904, 747 cmꢀ1
; elemental analysis calcd (%) for C12H14
(158.25): C 91.08, H 8.92; found: C 90.97, H 8.89.
(2E)-(1-Methylpenta-2,4-dienyl)benzene (4i): Compound 4h was ob-
tained from 1b and 2-methyl-2-phenylethanal (3i) in 70% yield as a mix-
ture of isomers (E/Z 95:5) following the general olefination procedure.
Comparison of the spectral data with the literature[24] confirmed the iden-
1
tity of compound 4i. H NMR (300 MHz, CDCl3): d=7.41–7.19 (m, 5H),
6.95–6.77 (m, 0.05H), 6.44 (ddd, J=16.8, 10.1, 10.0 Hz, 0.95H), 6.19 (dd,
J=15.2, 10.0 Hz, 1H), 5.99 (dd, J=15.2, 6.5 Hz, 0.95H), 5.67 (m, 0.05H),
5.24 (dd, J=16.8, 1.7 Hz, 1H), 5.11 (dd, J=10.1, 1.7 Hz, 1H), 4.14–3.98
(m, 0.05H), 3.70–3.55 (m, 0.95H), 1.52 ppm (d, J=7.2 Hz, 3H); 13C
NMR (75 MHz, CDCl3): d=145.6, 139.6, 137.3, 129.7, 128.0, 127.3, 126.3,
115.8, 42.3, 21.2 ppm; IR (CsI): n˜ =3064, 3027, 2928, 1802, 1602, 1492,
(E)-Pentadeca-1,3-diene (4b): Compound 4b was obtained from 1a and
dodecanal (3a) in 79% yield as a mixture of isomers (E/Z 94:6) follow-
ing the general olefination procedure. Comparison of the spectral data
with the literature[20] confirmed the identity of compound 4b. 1H NMR
(200 MHz, CDCl3): d=6.76–6.56 (m, 0.06H), 6.33 (ddd, J=17.0, 10.2,
10.1 Hz, 0.94H), 6.06 (dd, J=15.2, 10.2 Hz, 1H), 5.71 (dt, J=15.2,
6.9 Hz, 0.94H), 5.59–5.37 (m, 0.06H), 5.09 (dd, J=17.0, 1.8 Hz, 1H), 4.95
(dd, J=10.1, 1.8 Hz, 1H), 2.30–2.03 (m, 2H), 1.50–1.10 (m, 18H),
0.89 ppm (t, J=6.5 Hz, 3H); 13C NMR (50 MHz, CDCl3): d=137.4,
135.6, 133.1, 132.4, 130.8, 116.6, 114.5, 32.6, 31.9, 29.6, 29.5, 29.4, 29.2,
27.8, 22.7, 14.1 ppm; IR (CsI): n˜ =3086, 2925, 2854, 1793, 1652, 1603,
1451, 1008, 905, 699 cmꢀ1
; elemental analysis calcd (%) for C12H14
(158.25): C 91.08, H 8.92; found: C 91.05, H 8.90.
(E)-1-Phenyltridec-1-ene (7b): Compound 7b was obtained from 1d and
dodecanal (3b) in 63% yield as the pure E isomer following the general
olefination procedure. Comparison of the spectral data with the litera-
ture[25] confirmed the identity of compound 7b. 1H NMR (300 MHz,
CDCl3): d=7.39–7.18 (m, 5H), 6.41 (d, J=15.3 Hz, 1H), 6.25 (dt, J=
15.3, 6.9 Hz, 1H), 2.23 (q, J=6.9 Hz, 2H), 1.51–1.31 (m, 20H), 0.93 ppm
(t, J=6.2 Hz, 3H); 13C NMR (75 MHz, CDCl3): d=138.3, 131.6, 130.0,
128.8, 127.1, 126.2, 33.4, 32.3, 30.0, 29.9, 29.7, 29.6, 23.0, 14.5 ppm; IR
(CsI): n˜ =3081, 3060, 3025, 2955, 2924, 2853, 1494, 1466, 962, 742,
1465, 1378, 949, 722 cmꢀ1
(208.39): C 86.46, H 13.54; found: C 86.39, H 13.51.
(S,3E)-6,10-Dimethylundeca-1,3,9-triene (4c): Compound 4c was ob-
; elemental analysis calcd (%) for C15H28
ACHTREUNG
tained from 1a and (ꢀ)-1-citronellal (3c) in 72% yield as a mixture of
1
isomers (E/Z 97:3) following the general olefination procedure. H NMR
(200 MHz, CDCl3): d=6.34 (ddd, J=16.8, 10.2, 10.1 Hz, 1H), 6.05 (dd,
J=15.1, 10.2 Hz, 1H), 5.71 (dt, J=15.1, 7.6 Hz, 1H), 5.14–5.04 (m, 2H),
4.96 (dd, J=9.9, 1.7 Hz, 1H), 2.23–1.85 (m, 4H), 1.69 (s, 3H), 1.61 (s,
3H), 1.43–1.10 (m, 3H), 0.88 ppm (d, J=6.6 Hz, 3H); 13C NMR
(50 MHz, CDCl3): d=137.4, 134.0, 132.2, 131.1, 124.9, 114.5, 40.0, 36.8,
32.9, 25.7, 19.5, 17.6 ppm; IR (CsI): n˜ =3059, 2962, 2921, 1798, 1651,
1602, 1455, 1377, 1260, 1002, 897, 739 cmꢀ1; elemental analysis calcd (%)
for C13H22 (178.32): C 87.56, H 12.44; found: C 87.55, H 12.39.
691 cmꢀ1
.
(1E)-1,4-Diphenylbut-1-ene (7 f): Compound 7 f was obtained from 1d
and 3-phenylpropanal (3 f) in 60% yield as the pure E isomer following
the general olefination procedure. Comparison of the spectral data with
the literature[26] confirmed the identity of compound 7 f. 1H NMR
(200 MHz, CDCl3): d=7.51–7.27 (m, 10H), 6.57 (d, J=15.9 Hz, 1H),
6.40 (dt, J=15.9, 6.3 Hz, 1H), 2.94 (t, J=8.0 Hz, 2H), 2.73–2.62 ppm (m,
2H); 13C NMR (50 MHz, CDCl3): d=141.7, 137.7, 130.4, 130.0, 128.5,
128.4, 126.9, 126.0, 125.9, 35.9, 34.8 ppm; IR (film): n˜ =3058, 3025, 2933,
(3E)-Buta-1,3-dienylcyclohexane (4d): Compound 4d was obtained from
1a and cyclohexanecarbaldehyde (3d) in 41% yield as a mixture of iso-
mers (E/Z 93:7) following the general olefination procedure. Comparison
of the spectral data with the literature[21] confirmed the identity of com-
pound 4d. 1H NMR (200 MHz, CDCl3): d=6.78–6.59 (m, 0.07H), 6.31
(ddd, J=16.8, 10.2, 10.1 Hz, 0.93H), 6.03 (dd, J=15.2, 10.2 Hz, 1H), 5.66
(dd, J=15.2, 6.8 Hz, 0.93H), 5.40–5.22 (m, 0.07H), 5.10 (dd, J=16.8,
1.6 Hz, 1H), 4.97 (dd, J=10.1, 1.6 Hz, 1H), 2.03–1.94 (m, 1H), 1.86–1.56
(m, 5H), 1.37–1.00 ppm (m, 5H); 13C NMR (50 MHz, CDCl3): d=141.1,
139.3, 137.7, 132.6, 128.4, 114.6, 40.6, 36.7, 33.3, 32.8, 26.2, 26.0, 25.9 ppm;
IR (CsI): n˜ =2926, 2852, 1651, 1604, 1448, 1002, 950, 893 cmꢀ1; elemental
analysis calcd (%) for C10H16 (136.24): calcd C 88.16, H 11.84; found: C
88.05, H 11.82.
2850, 1945, 1872, 1748, 1651, 1599, 1494, 1457, 1433, 1071, 964, 739 cmꢀ1
;
elemental analysis calcd (%) for C16H16 (208.31): C 92.26, H 7.74; found:
C 92.05, H 7.72.
trans-Stilbene (7g): Compound 7g was obtained from 1d and benzalde-
hyde (3g) in 56% yield as the pure E isomer following the general olefi-
nation procedure. Comparison of the spectral data with the Aldrich data-
base confirmed the identity of compound 7g. 1H NMR (300 MHZ,
CDCl3): d=7.45 (d, J=7.3 Hz, 4H), 7.33–7.19 (m, 6H), 7.05 ppm (s,
2H); 13C NMR (75 MHz, CDCl3): d=137.3, 128.7, 128.6, 127.6,
126.5 ppm.
trans,trans-1,4-Diphenylbuta-1,3-diene (7h): Compound 7h was obtained
from 1d and cinnamaldehyde (3h) in 53% yield as the pure E isomer fol-
lowing the general olefination procedure. Comparison of the spectral
data with the Aldrich database confirmed the identity of compound 7g.
1H NMR (300 MHZ, CDCl3): d=7.46 (t, J=7.2 Hz, 4H), 7.35 (d, J=
7.2 Hz, 4H), 7.27–722 (m, 2H), 7.02–6.93 (m, 2H), 6.74–6.64 ppm (m,
2H); 13C NMR (75 MHz, CDCl3): d=137.5, 133.0, 129.4, 128.8, 127.7,
126.5 ppm.
(4E)-4-Hepta-4,6-dienyl-2,2-dimethyl-1,3-dioxolane (4e): Compound 4e
was obtained from 1a and 5,6-O-isopropylidenehexanal (3e)[22] in 85%
yield as a mixture of isomers (E/Z 94:6) following the general olefination
procedure. 1H NMR (200 MHz, CDCl3): d=6.75–6.54 (m, 0.06H), 6.28
(ddd, J=16.8, 10.3, 10.2 Hz, 0.94H), 6.03 (dd, J=15.2, 10.2 Hz, 1H), 5.70
(dt, J=15.2, =6.8 Hz, 0.94H), 5.55–5.37 (m, 0.06H), 5.09 (dd, J=16.8,
1.7 Hz, 1H), 4.94 (dd, J=10.3, 1.7 Hz, 1H), 4.13–4.00 (m, 2H), 3.55–3.46
(m, 1H), 2.20–2.05 (m, 2H), 1.69–1.30 ppm (m, 10H); 13C NMR
(50 MHz, CDCl3): d=137.1, 134.6, 132.1, 132.0, 131.4, 117.1, 114.9, 108.7,
75.9, 72.1, 69.4, 66.8, 33.1, 32.7, 32.4, 27.6, 26.9, 25.7, 25.3, 25.1 ppm; IR
(CsI): n˜ =3085, 2986, 2935, 2865, 1802, 1732, 1651, 1603, 1456, 1370, 1248,
1157, 859, 737 cmꢀ1; elemental analysis calcd (%) for C12H20O2 (196.29):
C 73.43, H 10.27; found: C 73.41, H 10.26.
[1] a) G. Wittig, G. Geissler, Justus Liebigs Ann. Chem. 1953, 580, 44–
57; b) G. Wittig, U. Schçllkopf, Chem. Ber. 1954, 87, 1318–1330.
[2] R. W. Hoffmann, Angew. Chem. 2001, 113, 1457–1462; Angew.
Chem. Int. Ed. 2001, 40, 1411–1416.
Chem. Eur. J. 2007, 13, 5433 – 5440
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5439

S. Meunier, C. Mioskowski et al.
[3] B. E. Maryanoff, A. B. Reitz, Chem. Rev. 1989, 89, 863–927.
[4] a) R. Tamura, K. Saegusa, M. Kakihana, D. Oda, J. Org. Chem.
1988, 53, 2723–2728; b) E. Vedejs, H. W. Fang, J. Org. Chem. 1984,
49, 210–212.
[5] R. Q. Liu, M. Schlosser, Synlett 1996, 1195–1196.
[6] a) R. Q. Liu, M. Schlosser, Synlett 1996, 1197–1198; b) Y. Wang,
F. G. West, Synthesis 2002, 99–103.
[7] Q. Wang, M. El Khoury, M. Schlosser, Chem. Eur. J. 2000, 6, 420–
426.
[8] H. S. He, C. W. Y. Chung, T. Y. S. But, P. H. Toy, Tetrahedron 2005,
61, 1385–1405.
[9] A. W. Johnson, J. O. Martin, Chem. Ind. 1965, 1726–1727.
[10] J. D. Hsi, M. Koreeda, J. Org. Chem. 1989, 54, 3229–3231.
[11] B. Boubia, C. Mioskowski, Tetrahedron Lett. 1989, 30, 5263–5266.
[12] B. Boubia, A. Mann, F. D. Bellamy, C. Mioskowski, Angew. Chem.
1990, 102, 1522–1523; Angew. Chem. Int. Ed. Engl. 1990, 29, 1454–
1456.
[15] P. M. Booth, C. M. J. Fox, S. V. Ley, J. Chem. Soc., Perkin Trans. 1
1987, 121–129.
[16] H. Daniel, M. Le Corre, Tetrahedron Lett. 1987, 28, 1165–1168.
[17] a) M. Schlosser, F. K. Christmann, A. Piskala, D. Coffinet, Synthesis
1971, 29–31; b) M. Schlosser, D. Coffinet, Synthesis 1971, 380–381;
c) E. J. Corey, H. Yamamoto, J. Am. Chem. Soc. 1970, 92, 226–228.
[18] With 1b the reactions were carried out at ꢀ788C.
[19] With 1d only 2.5 equivalents of base were used.
[20] H.-C. Shen, H.-L. Su, Y.-C. Hsueh, R.-S. Liu, Organometallics 2004,
23, 4332–4334.
[21] N. Chinkov, S. Majumpar, I. Marek, Synthesis 2004, 14, 2411–2417.
[22] K. L. McCaleb, R. L. Halcomb, Org. Lett. 2000, 2, 2631–2634.
[23] J. Uenishi, R. Kawahama, O. Yonemitsu, J. Tsuji, J. Org. Chem.
1996, 61, 5716.
[24] Y. Ikeda, H. Yamamoto, Bull. Chem. Soc. Jpn. 1986, 59, 657–658.
[25] A. Orita, A. Watanabe, H. Tsuchiya, J. Otera, Tetrahedron 1999, 55,
2889–2898.
[13] E. G. McKenna, B. J. Walker, Tetrahedron Lett. 1988, 29, 485–488.
[14] a) T. Kaufmann, H. Fischer, A. Woltermann, Angew. Chem. 1977,
89, 52–53; Angew. Chem. Int. Ed. Engl. 1977, 16, 53–54; b) T. Kauf-
mann, H. Fischer, A. Woltermann, Chem. Ber. 1982, 115, 645–653.
[26] P. L. Wylie, K. S. Prowse, M. A. Belill, J. Org. Chem. 1983, 48, 4022–
4025.
Received: October 30, 2006
Published online: February 9, 2007
5440
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 5433 – 5440